Project Summary This research will focus on immunotherapy and the tumor microenvironment in breast cancer, and particularly triple-negative breast cancer (TNBC), which is highly metastatic, has the worst prognosis among breast cancer subtypes, and is lacking effective therapies. Immunotherapy is changing the paradigm of cancer treatment, but in breast cancer the response rate to single agent immune checkpoint blockade is low, compared to more immunogenic cancers. A quantitative understanding of the complexity of the immune-cancer interactions is presently insufficient. The long-term goal of this project is to develop predictive, mechanistic clinically- and experimentally-based computational models of breast cancer, taking into account the immune-cancer interactions, and apply them to modeling cancer immunotherapy. The project will be a close collaboration between computational, clinical, and experimental researchers. We will formulate quantitative systems pharmacology (QSP) ordinary differential equation-based models comprising tumor (primary and metastasis), lymph nodes, and blood and peripheral compartments; we will also formulate spatio-temporal three- dimensional agent-based and hybrid tumor models that will describe tumor heterogeneity that is a hallmark of cancer. Transport of ligands and drugs will be modeled by partial differential equations. The data for these spatial models will be derived from our computational analysis of clinical pathology images where we will determine the spatial distributions of immune cells, such as CD8+ T cells, regulatory T cells, and myeloid- derived suppressor cells, and molecular markers such as PD-1, PD-L1, PD-L2, FoxP3, and LAG-3. The distributions will be used to parameterize and validate the models; part of these data will serve as the input to computational models and part for model validation. We will conduct state-of-the-art sensitivity analysis and uncertainty quantification. The computer codes will be reported in the form to share with the research community, to ensure reproducibility. The clinical data will be derived from several breast cancer immunotherapy clinical trials in which immune checkpoints CTLA-4, PD-1, and PD-L1 are targeted, in combination with immunomodulating agents, e.g. epigenetic. Clinical data will be supplemented with experimental data obtained from syngeneic mouse models with orthotopic triple-negative breast cancer tumors, with the experimental protocols mimicking the clinical trials. A variety of experimental methods will be used to provide a plethora of data for model parameterization and validation, including flow cytometry, immunofluorescence microscopy, protein arrays, and molecular biology. Additional immune checkpoints will be explored experimentally and computationally, such as OX40 and LAG-3. The research will contribute to a fundamental understanding of breast cancer biology, to the identification of potential biomarkers, and will aid in design and interpretation of clinical trials. The synergistic combination of computational, clinical, and experimental studies will provide significant insights into breast cancer immunotherapy.